Objective This study examined the associations between changes in common mental disorders (CMD) and subsequent diagnosis-specific sickness absence (SA) among midlife and ageing municipal employees.

Methods Data from the Helsinki Health Study phase I (2000–2002) and phase II (2007) surveys among employees of the City of Helsinki, Finland, were linked with prospective register data from the Social Insurance Institution of Finland on diagnosis-specific (mental, musculoskeletal, other causes) SA (n=3890). Associations between change in CMD (General Health Questionnaire 12) from phase I to phase II and the first SA event in 2007–2014 were analysed using Cox regression modelling. Sociodemographic, work and health-related covariates from phase I, and SA from the year preceding phase I were controlled for.

Results Having CMD at one or two time points, that is, favourable and unfavourable change in CMD and repeated CMD, were all associated with a higher risk of SA due to mental, musculoskeletal and other diagnoses compared with women and men with no CMD. Favourable change in CMD reduced the risk of SA when compared with repeated CMD. The strongest associations were observed for repeated CMD (HR range: 1.44 to 5.05), and for SA due to mental diagnoses (HR range: 1.15 to 5.05). The associations remained after adjusting for the covariates.

Conclusions Changing and repeated CMD increased the risk of SA due to mental, musculoskeletal and other diagnoses. CMD should be tackled to prevent SA and promote work-ability among ageing employees.